tulit, The Universal Legal Informatics Toolkit

1. Introduction

The tulit package provides utilities to work with legal data in a way that legal informatics practitioners can focus on adding value.

2. Getting started

Documentation is available at https://tulit-docs.readthedocs.io/en/latest/index.html

2.1 Installation

To install the tulit package, you can use the following command:

pip install tulit

or using poetry:

poetry add tulit

2.2 Basic usage

The tulit package main components are:

• a client to query and retrieve data from a variety of legal data sources. Currently the package supports the Cellar, LegiLux and Normattiva.
• a parser to convert legal documents from a variety of formats to a standardised json representation.

Retrieving legal documents

The tulit package provides a client to query and retrieve data from a variety of legal data sources. The following code snippet shows how to use the tulit package to retrieve a legal document from Cellar, given its CELEX number:

from tulit.download.cellar import CellarDownloader

client = CellarDownloader()
downloader = CellarDownloader(download_dir='./tests/data/formex', log_dir='./tests/logs')
with open('./tests/metadata/query_results/query_results.json', 'r') as f:
    results = json.loads(f.read())
documents = downloader.download(results, format='fmx4')

print(documents)

Parsing legal documents

The tulit parsers support exclusively legislative documents which were adopted in the following formats:

• Akoma Ntoso 3.0
• FORMEX 4
• XHTML originated from Cellar

The following code snippet shows how to use the tulit package to parse a legal document in Akoma Ntoso format:

from tulit.parsers.xml.akomantoso import AkomaNtosoParser

parser = AkomaNtosoParser()
    
file_to_parse = 'tests/data/akn/eu/32014L0092.akn'
parser.parse(file_to_parse)

# The various attributes of the parser can be accessed as follows
print(parser.preface)
print(parser.citations)
print(parser.recitals)
print(parser.chapters)
print(parser.articles)

A similar approach can be used to parse a legal document in FORMEX and XHTML format:

from tulit.parsers.xml.formex import FormexParser

formex_file = 'tests/data/formex/c008bcb6-e7ec-11ee-9ea8-01aa75ed71a1.0006.02/DOC_1/L_202400903EN.000101.fmx.xml'
parser = FormexParser()

parser.parse(formex_file)

from tulit.parsers.html.xhtml import HTMLParser

html_file = 'tests/data/html/c008bcb6-e7ec-11ee-9ea8-01aa75ed71a1.0006.03/DOC_1.html'

parser = HTMLParser()
parser.parse(html_file)

Use of existing standards and structured formats

The tulit package is designed to work with existing standards and structured formats in the legal informatics domain. The following are some of the standards and formats that the package is designed to work with:

• LegalDocML (Akoma Ntoso)
• FORMEX

Further standards and formats will be added in the future such as:

• LegalHTML
• NormeInRete

Acknowledgements

The tulit package has been inspired by a series of existing resources and builds upon some of their architectures and workflows. We would like to acknowledge their work and thank them for their contributions to the legal informatics community.

• The eu_corpus_compiler repository by Selja Seppala concerning the methods used to query the CELLAR SPARQL API and WEB APIs
• The sortis project results from the European Commission
• The EURLEX package by step 21
• The eurlex package by kevin91nl
• The extraction_libraries by the Maastricht Law and Tech Lab
• The closer library by the Maastricht Law and Tech Lab